Prior authorisation denial cost calculator

How each figure is built

Monthly denials are volume multiplied by the denial rate. Appeals are the share of those denials your team actually works. Labour hours are appeals multiplied by minutes per appeal, divided by sixty, and labour cost applies the fully loaded hourly rate. Recovered revenue is the appeals that overturn, multiplied by average reimbursement. Abandoned revenue applies the same overturn rate to the denials nobody appealed, on the assumption that an unworked denial would have behaved like a worked one. Cost per successful appeal is annual labour cost divided by the number of appeals that overturn.

Assumptions and limits

Every denial is treated as identical. In a real department they are not: a missing clinical note and a benefit exclusion cost different minutes and overturn at different rates, and the average hides that a small number of complex cases consume most of the queue. If your mix is lopsided, run the model twice with different inputs rather than once with a blend.

The abandoned revenue figure is the softest number on this page. It assumes denials that were never appealed would have overturned at the same rate as those that were, and that is generous. Teams triage, and what gets dropped is usually what looked least winnable, so the true recoverable share is lower. Read it as an upper bound on what the queue is leaving behind, not as a forecast.

Reimbursement is treated as revenue rather than margin, so it overstates the economic value of a recovered case by the cost of delivering the care. The model excludes the downstream cost of a denial that sticks: the write off, the patient balance that will not be collected, the rework in scheduling when a case is cancelled, and the clinical time already spent. It excludes payer-side timelines, so an appeal that wins after the filing window is counted as a win. It excludes peer-to-peer physician time, which is the most expensive hour in the process and rarely lands in the revenue cycle budget. It excludes any effect on the patient, who is waiting throughout.

Nothing here is a clinical, safety or regulatory claim, and nothing here is a statement about any particular product.

What this does not tell you

It cannot tell you whether an automation vendor's denial numbers are evidenced. That is a separate question with a separate method, and the scoring we use for it is published at how we score vendors, including what separates an asserted figure from a verified one.

If you need the full instrument, with denial reason categories, a leak pricing model and the scorecard for judging a denial-prediction vendor, that lives in the Revenue Cycle AI Toolkit.